Abstract
Reduced iron was abundant in Earth's surface environments before their oxygenation, so iron oxidation could have been a common metabolism on the early Earth. Consequently, modern microbial iron oxidation is sometimes seen as a holdover from an earlier biosphere, but the continuity of involved lineages or the metabolic process itself has not been verified. Modern neutrophilic iron oxidizers use cytochrome-porin Cyc2 as the initial electron acceptor in iron oxidation. With the protein as a proxy for the metabolism, we performed a phylogenetic analysis of Cyc2 to understand the evolutionary history of this microbial iron oxidation pathway. In addition to known iron oxidizers, we identified Cyc2 orthologs in gammaproteobacterial endosymbionts of lucinid bivalves. These bivalves have a robust fossil record and rely on seagrass meadows that only appear in the Cretaceous, providing a valuable time calibration in the evolutionary history of Cyc2. Our molecular clock analysis shows that extant sampled Cyc2 diversity has surprisingly recent common ancestry, and iron oxidation metabolisms in Gallionellaceae, Zetaproteobacteria, and photoferrotrophic Chlorobi likely originated in the Neoproterozoic or the Phanerozoic via multiple transfer events. The groups responsible for microbial iron oxidation have thus changed over Earth history, possibly reflecting the instability of niches with sufficient reduced iron. We note that frequent transfer and changing taxonomic distribution may be a general pattern for traits which are selected sporadically across space and time. Based on iron metabolism and other processes, we explore this concept of a trait's "clade fidelity" (or lack thereof) and establish its evolutionary importance.IMPORTANCEBacteria can oxidize iron to produce energy. As there was plenty of reduced iron available on the early Earth and there is only a little today, it was sometimes thought that bacteria that oxidize iron today are a small remnant of a larger group that used to do it. We studied the evolutionary history of the iron oxidation pathway that modern bacteria use, and we found that they developed that pathway relatively recently: whatever did it in the past is no longer around today. It would probably be hard for any group of organisms to keep doing iron oxidation over billions of years since iron availability is so variable: they are likely to go extinct or lose this ability at some point. We suggest this as a general trend in evolution that traits which are only sporadically useful are commonly lost-and then re-invented or re-distributed-or the trait will go extinct.